Distributed systems in practice, in theory

Confidential + ProprietaryConfidential + Proprietary

Distributed Systems in Practice, in TheoryAysylu GreenbergMarch 8th, 2016

Confidential + Proprietary



Aysylu Greenberg

@aysylu22


Towards Distributed Build System




http://opencv.org/



http://opencv.org/



http://opencv.org/



http://opencv.org/



http://opencv.org/



http://opencv.org/



http://opencv.org/



http://opencv.org/



http://opencv.org/



http://opencv.org/



http://opencv.org/



http://opencv.org/



http://opencv.org/



http://opencv.org/



http://opencv.org/




Papers We Love SF


Aysylu Greenberg

@aysylu22

Confidential + Proprietarymeetup.com/Papers-We-Love-London


Today

● Staged Event-Driven Architecture


Today

● Staged Event-Driven Architecture● Leases


Today

● Staged Event-Driven Architecture● Leases● Inaccurate Computations


Computer Science ResearchInDistributed Systems Industry


Operating systems research





Concurrency



Concurrency

Concurrency primitives: mutex & semaphore



Concurrency

Concurrency primitives: mutex & semaphore

Processes execute at different speeds


Time in distributed systems





Pipelining



Pipelining


Internet


Internet

Distributed consensus


Internet



Internet



Internet


Paxos


Reconsider large systems


Reconsider large systems

Platform as a service


CS Research is Timeless

Inform decisions

Mitigate technical risk


Staged Event Driven Architecture&Deep Pipelines

2001


Hardware Pipelines


Data Pipelines

https://en.wikipedia.org/wiki/Graphics_pipeline


Search Indexing Pipelines























Staged Event Driven Architecture


Single-machine pipeline

generalizes to distributed pipelines




● Dynamic resource controllers



● Dynamic resource controllers

+-



● Dynamic resource controllers○ automatic tuning

+-




■ thread pool sizing+

-




■ thread pool sizing■ event batching +

-


+-



■ thread pool sizing■ event batching

○ load shedding via queue


+-




○ load shedding via queue■ backpressure


+-




○ load shedding via queue■ backpressure■ threshold, filter, reorder, aggregate


Leasesas Heart Beat in Distributed Systems 1989



Leases

● Distributed locking


Leases

● Distributed locking● Lease term tradeoffs

○ short


Leases


○ short■ delay from client and server failures

minimized■ reduced false write-sharing


Leases


○ short vs long


Leases


○ short vs long■ more efficient for frequently accessed

data■ minimized lease extension overhead on

server and client


Leases


○ short vs long● Use of leases in modern applications

○ Leader election TTL (in etcd)


Leases


○ short vs long● Use of leases in modern applications

○ Leader election TTL (in etcd)○ Liveness detection



Do the build, s'il vous plaît!

Build System

Dita Roque-Gourary



Build System

OK/D'Accord!

Dita Roque-Gourary



Build System

OK/D'Accord!

Waiting for the results, merci.

Dita Roque-Gourary



Build System

OK/D'Accord!


Build is in progress

Dita Roque-Gourary



Build System

OK/D'Accord!



Dita Roque-Gourary




Build System

OK/D'Accord!



Build is finished/fini.Dita Roque-Gourary



Leases in Build System


Inaccurate Computations&Serving Search Results


From Accurate to"Good Enough"







Inaccuracy for Resilience

1. Task decomposition



1. Task decomposition2. Baseline for correctness



1. Task decomposition2. Baseline for correctness3. Criticality Testing



1. Task decomposition2. Baseline for correctness3. Criticality Testing4. Distortion and timing models

Confidential + ProprietaryJeff Dean "Building Software Systems at Google and Lessons Learned", Stanford, 2010

Confidential + ProprietaryJeff Dean "Building Software Systems at Google and Lessons Learned", Stanford, 2010




Robust & scaleable pipelines


Robust & scaleable pipelinesLeases for sharing & heartbeat


Robust & scaleable pipelinesLeases for sharing & heartbeat

Trade off inaccuracy for resilience & performance


Robust, scaleable pipelinesLeases for sharing & heartbeat

Trade off inaccurate for resilient

CS research is timeless:

use it to mitigate risk


ThanksInes SombraDavid GreenbergAlex HutchesonScott ZawalskiKaran Parikh


References● T. Wurthinger, C. Wimmer et al. "One VM to Rule Them All"● M. Rinard "Probabilistic Accuracy Bounds for Fault-Tolerant

Computations that Discard Tasks"● F. Corbato, M. Daggett, R. Daley "An Experimental Time-

Sharing System"● E. Dijkstra "Cooperating Sequential Processes"● L. Lamport "Time, Clocks, and the Ordering of Events in a

Distributed System"


References● B. Oki, B. Liskov "Viewstamped Replication: A New Primary Copy

Method to Support Highly-Available Distributed Systems"● L. Lamport "The Part-Time Parliament"● M. Welsh, D. Culler, E. Brewer "SEDA: An Architecture for Well-

Conditioned, Scalable Internet Services"● C. Gray, D. Cheriton "Leases: An Efficient Fault-Tolerant

Mechanism for Distributed File Cache Consistency"● S. Agarwal, B. Mozafari et al. "BlinkDB: Queries with Bounded

Errors and Bounded Response Times on Very Large Data"


Should I read papers?

YES